PID parameter debugging device based on AI calculation

Abstract

The invention particularly relates to a PID parameter debugging device based on AI calculation, and belongs to the technical field of PID parameter debugging. The device comprises an execution module used for preliminarily setting PID parameters according to a control target of a PID controller, setting output limiting conditions of the PID controller, and inputting dynamic response data of a controlled object in a time window; and the AI controller is used for carrying out optimization iteration on the PID parameters and outputting the optimal PID parameters. According to the method, iterative optimization is carried out on the PID parameters through AI calculation, the robustness performance of the PID parameters is improved, the requirement of nuclear power scenes for robustness performance indexes is met, and meanwhile the manpower cost and the time cost of PID parameter debugging are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of PID parameter debugging, in particular to a PID parameter debugging device based on AI calculation, which is suitable for nuclear power scenarios with strict requirements for robustness performance. BACKGROUND

[0002] When the PID controller debugging work is carried out, the PID parameters are initially drafted by the debugging person in charge, and then the trial-and-error of the PID parameters is carried out many times according to the debugging experience and the change of the response curve in the debugging process, and the best PID parameter with the best robustness is selected manually. This method consumes a lot of manpower and time cost, and the manually selected parameters are difficult to cover complex working conditions such as noise interference, and cannot meet the requirements of nuclear power scenarios for robustness performance. SUMMARY

[0003] The purpose of the present application is to provide a PID parameter debugging device based on AI calculation, which iteratively optimizes the PID parameters through AI calculation, improves the robustness performance of the PID parameters, meets the requirements of nuclear power scenarios for robustness performance indicators, and reduces the manpower and time cost of PID parameter debugging.

[0004] In order to achieve the above purpose, the present application provides the following technical scheme:

[0005] A PID parameter debugging device based on AI calculation comprises:

[0006] An execution module is configured to initially set the PID parameters according to the control target of the PID controller, and set the output limit condition of the PID controller and the dynamic response data of the controlled object within the input time window.

[0007] An AI controller is configured to iteratively optimize the PID parameters and output the optimal PID parameters.

[0008] As one of the implementable ways, the PID parameters include proportional gain K p , differential gain K d , integral time T i and differential time T d ; the output limit condition of the PID controller includes output upper limit, output lower limit, output value change rate, dead zone and action direction; the dynamic response data of the controlled object within the time window includes the set value of the controlled object, the measured value of the controlled object, the output value of the actuator and the actual feedback value of the actuator.

[0009] As one of the implementable modes, a calculation confirmation button is arranged on the execution module; after the execution module preliminarily sets the PID parameters according to the control target of the PID controller, sets the output limit condition of the PID controller and the dynamic response data of the controlled object in the input time window, the calculation confirmation button is pressed to trigger the AI controller to optimize and iterate the PID parameters and output the optimal PID parameters;

[0010] The AI controller iteratively optimizes the PID parameters and outputs the optimal PID parameters, including the following steps:

[0011] Step 1: The AI controller establishes a PID controller transfer function according to the PID parameters, establishes a controlled object transfer function according to the dynamic response data of the controlled object in the time window, establishes a closed-loop system transfer function according to the PID controller transfer function and the controlled object transfer function, and establishes a closed-loop system characteristic equation according to the closed-loop system transfer function;

[0012] Step 2: The AI controller constructs a Routh table according to the closed-loop system characteristic equation, judges whether the closed-loop system is stable according to the Routh table and a preset Routh criterion, adjusts the PID parameters if the closed-loop system is unstable and returns to step 1, and enters step 3 if the closed-loop system is stable;

[0013] Step 3: The AI controller evaluates whether the unit step calculation performance index of the closed-loop system meets the preset requirement, adjusts the PID parameters if the preset requirement is not met and returns to step 1, and outputs the PID parameters at this time as the optimized PID parameters if the preset requirement is met.

[0014] As one of the implementable modes, the Routh criterion is that all elements in the first column of the Routh table are > 0, and the closed-loop system is determined to be stable; if there is an element ≤ 0 in the first column of the Routh table, the closed-loop system is determined to be unstable;

[0015] If the closed-loop system is unstable, the AI controller adjusts the PID parameters according to the sign change of the elements in the first column of the Routh table, specifically:

[0016] If a negative sign appears in the highest order row of the elements in the first column of the Routh table, the AI controller decreases K p according to a preset decrement;

[0017] If a negative sign appears in the middle row of the elements in the first column of the Routh table, the AI controller increases T d according to a preset increment and increases T i according to the preset increment;

[0018] If a negative sign appears in the constant row of the elements in the first column of the Routh table, the AI controller increases T i according to a preset increment or decreases Kp ;

[0019] If the first column element of the Louth table appears zero element, the AI controller increases T according to the preset increment d and decreases K according to the preset decrement p .

[0020] As one of the implementable ways, the unit step calculation performance index of the closed loop system includes the rise time T r , the regulation time T s , and the overshoot σ %.

[0021] The AI controller synchronously records the change of the response curve in the PID parameter iterative optimization process, evaluates whether the unit step calculation performance index of the closed loop system meets the preset requirements, and if not, adjusts the PID parameters, specifically:

[0022] If the overshoot of the closed loop system does not meet the preset requirements, the AI controller decreases K according to the preset decrement p or increases K according to the preset increment d .

[0023] If the regulation time of the closed loop system does not meet the preset requirements, the AI controller increases K according to the preset increment p .

[0024] As one of the implementable ways, the AI controller establishes the closed loop system transfer function T(s) according to the PID controller transfer function G(s) and the controlled object transfer function G0(s).

[0025] The AI controller establishes the closed loop system characteristic equation D(s) according to the closed loop system transfer function T(s), including the following steps:

[0026] Let the denominator polynomial of the closed loop system transfer function T(s) be 0, and after dividing, extract the numerator polynomial to convert it into a standard polynomial as the closed loop system characteristic equation D(s) = 0; the roots of D(s) = 0 are the closed loop system poles, which determine the stability of the closed loop system.

[0027] As one of the implementable ways, the closed loop system is a unit negative feedback structure, and the closed loop system transfer function T(s) is:

[0028]

[0029] The PID controller has a derivative filter; the PID controller transfer function G(s) is:

[0030]

[0031] The controlled object transfer function G0(s) is:

[0032]

[0033] Closed-loop system characteristic equation D(s) = c k ×s k +c k-1 ×s k-1 +...+c1×s+c0 = 0.

[0034] where K p is a proportional gain, K d is a derivative gain; T i is an integral time; T d is a derivative time; s is a complex variable, b0, b1,..., b m-1 , b m and a0, a1,..., a n-1 , a n are coefficients;

[0035] c i is a function of K p , K d , T i , T d , b y , a x ; i = 0, 1,..., k-1, k; y = 0, 1,...,

[0036] m-1, m; x = 0, 1,..., n-1, n.

[0037] As one of the implementable ways, the AI controller constructs the Routh table according to the closed-loop system characteristic equation, including the following steps:

[0038] Step 1, ensure that c k > 0; if c k < 0, multiply all the coefficients of the closed-loop system characteristic equation by -1;

[0039] Step 2, construct the Routh table, the first row elements of the Routh table are the even-order coefficients starting from the highest-order coefficient c k , equal to c k , c k-2 , c k-4 ,...;

[0040] The second row elements of the Routh table are the odd-order coefficients starting from the second highest-order coefficient c k-1 , equal to c k-1 , c k-3 , c k-5 ,...;

[0041] The third row of the Routh table starts, and each element of each row is equal to (the first item of the previous row multiplied by the next column item of the two previous rows minus the next column item of the previous row multiplied by the first item of the two previous rows) / the first item of the previous row;

[0042] If the first item of a row is 0 but the remaining items are not 0, continue to calculate after replacing 0 with a very small positive number ∈; if all items of a row are 0, construct an auxiliary equation with the previous row and derive, and fill the derivative coefficients into a new row.

[0043] The beneficial technical effects of the present application are:

[0044] The PID parameter debugging device based on AI calculation provided by the present application iteratively optimizes the PID parameters through AI calculation, greatly improves the robustness performance of the PID parameters after iterative optimization, meets the requirements of nuclear power scene on robustness performance, and greatly reduces the labor cost and time cost of PID parameter debugging. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 The structure schematic diagram of one embodiment of the PID parameter debugging device based on AI calculation of the present application. DETAILED DESCRIPTION

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application; the description and claims of this application as well as the above abstract are intended to cover any and all adaptations or variations of the present application including any and all equivalents thereof.

[0047] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase "in an embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another.

[0048] The technical solutions of the present application are described clearly and completely below in combination with the drawings and specific embodiments.

[0049] The closed-loop system is composed of a PID controller, an actuator, a controlled object and a feedback loop, forming a closed-loop control; the feedback loop feeds back an error signal to the PID controller; the PID controller generates a control signal according to the error signal to drive the actuator to act so that the output of the controlled object gradually approaches the set value.

[0050] The performance of the PID controller is highly dependent on the tuning of the PID parameters, which is an engineering practice process based on the dynamic characteristics of the specific controlled object, and the goal is to achieve the best balance between stability, response speed, overshoot and noise immunity.

[0051] Referring to Figure 1 , the embodiment provides a PID parameter debugging device based on AI calculation, comprising:

[0052] The execution module is configured to preliminarily set the PID parameters according to the control target of the PID controller, and set the output limit condition of the PID controller and the dynamic response data of the controlled object in the input time window.

[0053] The AI controller is configured to iteratively optimize the PID parameters and output the optimal PID parameters.

[0054] The control target of the PID controller is to make the controlled object reach the set value through feedback adjustment. The controlled object can be temperature, pressure or flow.

[0055] In the embodiment, as one of the implementable modes, the PID parameters include proportional gain K p , derivative gain K d , integral time T i and derivative time T d ; the output limit condition of the PID controller includes output upper limit, output lower limit, output value change rate, dead zone and action direction; and the dynamic response data of the controlled object in the time window includes the set value of the controlled object, the measured value of the controlled object, the output value of the actuator and the actual feedback value of the actuator.

[0056] The shorter the interval in the time window, the higher the robustness performance index of the PID parameters calculated and debugged by the AI controller. The longer the time window, the higher the robustness performance index of the PID parameters calculated and debugged by the AI controller.

[0057] In the embodiment, as one of the implementable modes, a calculation confirmation button is arranged on the execution module; after the execution module preliminarily sets the PID parameters according to the control target of the PID controller, and sets the output limit condition of the PID controller and the dynamic response data of the controlled object in the input time window, the calculation confirmation button is pressed to trigger the AI controller to iteratively optimize the PID parameters and output the optimal PID parameters.

[0058] The AI controller iteratively optimizes the PID parameters and outputs the optimal PID parameters, including the following steps:

[0059] Step 1, the AI controller establishes a PID controller transfer function according to the PID parameters, establishes a controlled object transfer function according to the dynamic response data of the controlled object within the time window, establishes a closed-loop system transfer function according to the PID controller transfer function and the controlled object transfer function, and establishes a closed-loop system characteristic equation according to the closed-loop system transfer function;

[0060] Step 2, the AI controller constructs a Routh table according to the closed-loop system characteristic equation, judges whether the closed-loop system is stable according to the Routh table and a preset Routh criterion, if the closed-loop system is not stable, adjusts the PID parameters and returns to step 1, if the closed-loop system is stable, enters step 3;

[0061] Step 3, the AI controller evaluates whether the unit step calculation performance index of the closed-loop system meets the preset requirements, if not, adjusts the PID parameters and returns to step 1, if yes, outputs the PID parameters at this time as the optimized PID parameters.

[0062] The AI controller repeatedly steps 1-3 to continuously iteratively optimize the PID parameters until the closed-loop system is stable and the unit step calculation performance index meets the preset requirements, and the AI controller outputs the PID parameters at this time as the optimized PID parameters.

[0063] The Routh criterion is a classical algebraic method for judging the stability of a linear time-invariant system, the core of which is to construct a Routh table through the coefficients of a characteristic equation, and to judge the stability of the system according to the sign changes of the elements in the first column of the table.

[0064] In this embodiment, as one of the implementable ways, the Routh criterion is: if all elements in the first column of the Routh table are > 0, it is determined that the closed-loop system is stable, if there is an element ≤ 0 in the first column of the Routh table, it is determined that the closed-loop system is not stable;

[0065] If the closed-loop system is not stable, the AI controller adjusts the PID parameters according to the sign changes of the elements in the first column of the Routh table, specifically:

[0066] If a negative sign appears in the highest order row of the elements in the first column of the Routh table, the AI controller decreases K p according to a preset decrement;

[0067] If a negative sign appears in the middle row of the elements in the first column of the Routh table, the AI controller increases T d according to a preset increment and increases T i according to a preset increment;

[0068] If a negative sign appears in the constant row of the elements in the first column of the Routh table, the AI controller increases T i according to a preset increment or decreases K p according to a preset decrement;

[0069] If the first column element of the Louth table appears a zero element, the AI controller increases T by a preset increment d and decreases K by a preset decrement p .

[0070] In this embodiment, as one of the implementable ways, the unit step calculation performance indicators of the closed-loop system include the rise time T r , the regulation time T s , and the overshoot σ%;

[0071] The AI controller synchronously records the changes of the response curve in the PID parameter iterative optimization process, evaluates whether the unit step calculation performance indicators of the closed-loop system meet the preset requirements, and if not, adjusts the PID parameters, specifically:

[0072] If the overshoot of the closed-loop system does not meet the preset requirements, the AI controller decreases K by a preset decrement p or increases K by a preset increment d .

[0073] If the overshoot time of the closed-loop system does not meet the preset requirements, the AI controller increases K by a preset increment p .

[0074] The overshoot σ% = ((peak value of the response curve - steady-state value of the response) / steady-state value of the response) × 100%; the rise time T r = the time required for the response to rise from 10% steady-state value to 90% steady-state value; the regulation time T s = the shortest time for the response to enter and permanently remain in the steady-state value allowable error band.

[0075] In this embodiment, as one of the implementable ways, the AI controller establishes the closed-loop system transfer function T(s) according to the PID controller transfer function G(s) and the controlled object transfer function G0(s);

[0076] The AI controller establishes the closed-loop system characteristic equation D(s) according to the closed-loop system transfer function T(s), including the following steps:

[0077] Let the denominator polynomial of the closed-loop system transfer function T(s) be 0, and after dividing, extract the numerator polynomial to convert it into a standard polynomial as the closed-loop system characteristic equation D(s) = 0; the roots of D(s) = 0 are the closed-loop system poles, which determine the stability of the closed-loop system.

[0078] In this embodiment, as one of the implementable ways, the closed-loop system is a unit negative feedback structure, and the closed-loop system transfer function T(s) is:

[0079]

[0080] PID controller with derivative filter; PID controller transfer function G(s) is:

[0081]

[0082] Controlled object transfer function G0(s) is:

[0083]

[0084] Closed-loop system characteristic equation D(s) = c k ×s k +c k-1 ×s k-1 +...+c1×s+c0=0;

[0085] Wherein, K p is proportional gain, K d is derivative gain; T i is integral time; T d is derivative time; s is complex variable, b0, b1,..., b m-1 , b m and a0, a1,..., a n-1 , a n are coefficients;

[0086] c i is a function of K p , K d , T i , T d , b y , a x ; i = 0, 1,..., k-1, k; y = 0, 1,..., m-1, m; x = 0, 1,..., n-1, n.

[0087]

[0088] G(s) is the PID controller transfer function, which describes the behavior of the PID controller in the complex frequency domain (s domain) to calculate the control signal u(t) according to the current error signal e(t); s is the Laplace transform complex variable.

[0089] K p ×1 is the proportional term, which provides proportional action proportional to the size of the current error; the proportional term responds quickly, but cannot eliminate steady-state error.

[0090] is the integral term, which provides integral action proportional to the accumulation of error over time; the core role of the integral term is to eliminate steady-state error, but may introduce response lag or overshoot, and even lead to instability. ​

[0091] The derivative term with filter provides a derivative action proportional to the rate of change of error; where, K p × T d is the ideal derivative term, is the filter; the ideal derivative term is extremely sensitive to high frequency noise in the measurement signal, which can amplify the noise, causing the PID controller output to oscillate wildly, damaging the actuator or making the closed loop system unstable; adding a filter can attenuate the high frequency noise, making the derivative action more practical and robust; the core role of the derivative term with filter is to predict the future error trend, provide damping effect, reduce overshoot, speed up response, and improve stability; the derivative term with filter makes it insensitive to noise.

[0092] The PID controller generates a control signal by a linear combination of the proportional term, the integral term, and the derivative term with filter.

[0093] U(s) is the Laplace transform of the control signal u(t) output by the PID controller, which drives the actuator to act to make the controlled object gradually approach the set value.

[0094] E(s) is the Laplace transform of the error signal e(t) input to the PID controller by the feedback loop, e(t) = r(t) - y(t), where r(t) is the set value and y(t) is the measured value.

[0095] K p is the proportional gain, representing the strength of the proportional action, determining the reaction strength of the PID controller to the current error size; K p is larger, the faster the response speed of the PID controller and the smaller the steady-state error, but it may lead to a decrease in stability; K p is smaller, the slower the response speed of the PID controller and the stronger the stability, but the steady-state error may increase; K p is the unit of the controller output unit / process variable unit.

[0096] K d is the derivative gain, representing the strength of the derivative filter, determining the suppression ability of the PID controller to high frequency measurement noise; K d is larger, the weaker the derivative filter action, the closer the derivative action to the ideal, the faster the response and the stronger the predictability, but it is more sensitive to noise; K d is smaller, the stronger the derivative filter action, the stronger the robustness of the derivative action to high frequency noise, but the real derivative action is smoothed in the high frequency band, and the response is slower; K d is dimensionless, with a value range of 5 to 20.

[0097] T iT i The greater, the weaker the integral effect, the slower the speed of eliminating steady-state error, and the steady-state error may persist, but the overshoot and oscillation will decrease, and the stability will increase;T i The smaller, the stronger the integral effect, the faster the speed of eliminating steady-state error, but it may lead to an increase in overshoot, exacerbate oscillation, and decrease stability;T i The unit is time.

[0098] T d T d The greater, the stronger the differential effect, the stronger the effect of suppressing overshoot, reducing oscillation, improving stability, and accelerating response speed; but too large T d may cause the PID controller output to fluctuate violently, damaging the actuator or making the closed-loop system unstable;T d The smaller, the weaker the differential effect, and the effect of suppressing overshoot and oscillation will be worse;T d The unit is time.

[0099] In this embodiment, as one of the implementable ways, the AI controller constructs the Routh table according to the characteristic equation of the closed-loop system, including the following steps:

[0100] Step 1, ensure that c k > 0; if c k < 0, multiply all coefficients of the characteristic equation of the closed-loop system by -1;

[0101] Step 2, construct the Routh table, the first row elements of the Routh table are even-numbered coefficients starting from the highest power coefficient c k , equal to c k , c k-2 , c k-4 ,...;

[0102] The second row elements of the Routh table are odd-numbered coefficients starting from the next highest power coefficient c k-1 , equal to c k-1 , c k-3 , c k-5 ,...;

[0103] The third row starts, and each row element = (the first item of the previous row x the next column item of the two previous rows - the next column item of the previous row x the first item of the two previous rows) / the first item of the previous row;

[0104] If the first item of a certain row is 0 but the remaining items are not zero, replace 0 with a very small positive number ∈ and continue to calculate; if all items in a certain row are 0, use the previous row to construct an auxiliary equation and take the derivative, and fill the derivative coefficients into the new row.

[0105] The above embodiments only express several implementation manners of the present application, which are described in a more specific and detailed manner, but cannot be understood as a limitation on the patent scope of the present application. It should be noted that, for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, which all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A PID parameter debugging device based on AI calculation, characterized in that: include: The execution module is used to preliminarily set the PID parameters according to the control target of the PID controller, and at the same time set the output limit conditions of the PID controller and the dynamic response data of the controlled object within the input time window; The AI ​​controller is used to optimize and iterate the PID parameters and output the optimal PID parameters.

2. The PID parameter debugging device based on AI calculation according to claim 1, characterized in that: PID parameters include proportional gain, differential gain, integral time and differential time; the output limit conditions of the PID controller include output upper limit, output lower limit, output value change rate, dead zone and action direction; the dynamic response data of the controlled object within the time window includes the set value of the controlled object, the measured value of the controlled object, the output value of the actuator and the actual feedback value of the actuator.

3. The PID parameter debugging device based on AI calculation according to claim 1 is characterized in that: A calculation confirmation button is provided on the execution module; after the execution module preliminarily sets the PID parameters according to the control target of the PID controller, and at the same time sets the output limit conditions of the PID controller and the dynamic response data of the controlled object within the input time window, press the calculation confirmation button to trigger the AI ​​controller to optimize and iterate the PID parameters and output the optimal PID parameters.

4. The PID parameter debugging device based on AI calculation according to claim 1, characterized in that: The AI ​​controller iteratively optimizes the PID parameters and outputs the optimal PID parameters, including the following steps: Step 1: The AI ​​controller establishes the PID controller transfer function based on the PID parameters; establishes the controlled object transfer function based on the dynamic response data of the controlled object within the time window; establishes the closed-loop system transfer function based on the PID controller transfer function and the controlled object transfer function; establishes the closed-loop system characteristic equation based on the closed-loop system transfer function; Step 2: The AI ​​controller constructs a Routh table based on the closed-loop system characteristic equation. Based on the Routh table and the preset Routh criterion, it determines whether the closed-loop system is stable. If the closed-loop system is unstable, the PID parameters are adjusted and the process returns to step 1. If the closed-loop system is stable, the process proceeds to step 3. Step 3: The AI ​​controller evaluates whether the unit step calculation performance index of the closed-loop system meets the preset requirements; if it does not meet the preset requirements, the PID parameters are adjusted and the process returns to step 1; if it meets the preset requirements, the PID parameters at this time are output as the optimized PID parameters.

5. The PID parameter debugging device based on AI calculation according to claim 4 is characterized in that: The Routh criterion is: if all elements in the first column of the Routh table are > 0, the closed-loop system is considered stable; if there is an element ≤ 0 in the first column of the Routh table, the closed-loop system is considered unstable.

6. The PID parameter debugging device based on AI calculation according to claim 4 is characterized in that: If the closed-loop system is unstable, the AI ​​controller adjusts the PID parameters according to the sign change of the elements in the first column of the Routh table. Specifically: If a negative sign appears in the highest-order row of the first column of the Routh table, the AI ​​controller will reduce K by the preset amount. p ; If a negative sign appears in the middle row of the first column of the Rolls table, the AI ​​controller increases T by the preset increment. d And increase T according to the preset increase i ; If a negative sign appears in the constant row of the first column of the Routh table, the AI ​​controller increases T by the preset increment. i Or reduce K by the preset reduction p ; If there is a zero element in the first column of the Routh table, the AI ​​controller increases T by the preset increment. d And reduce K according to the preset reduction p .

7. The PID parameter debugging device based on AI calculation according to claim 4 is characterized in that: Unit step calculation performance indicators of closed-loop systems include rise time, settling time and overshoot; The AI ​​controller synchronously records the changes in the response curve during the iterative optimization of the PID parameters, and evaluates whether the unit step calculation performance indicators of the closed-loop system meet the preset requirements. If the preset requirements are not met, the PID parameters are adjusted as follows: If the overshoot of the closed-loop system does not meet the preset requirements, the AI ​​controller will reduce K by the preset amount. p Or increase K by the preset increment d ; If the overshoot time of the closed-loop system does not meet the preset requirements, the AI ​​controller increases K by the preset increment. p .

8. The PID parameter debugging device based on AI calculation according to claim 4 is characterized in that: The AI ​​controller establishes the closed-loop system transfer function T(s) based on the PID controller transfer function G(s) and the controlled object transfer function G0(s); The AI ​​controller establishes the closed-loop system characteristic equation D(s) based on the closed-loop system transfer function T(s), which includes the following steps: Let the denominator polynomial of the closed-loop system transfer function T(s) be 0, and after common denominator, extract the numerator polynomial and convert it into a standard polynomial, which is used as the characteristic equation of the closed-loop system D(s) = 0; the root of D(s) = 0 is the pole of the closed-loop system and determines the stability of the closed-loop system.

9. The PID parameter debugging device based on AI calculation according to claim 8, characterized in that: The closed-loop system is a unit negative feedback structure, and the closed-loop system transfer function T(s) is: PID controller with differential filter; PID controller transfer function G(s) is: The transfer function of the controlled object G0(s) is: Closed-loop system characteristic equation D(s)=c k ×s k +c k-1 ×s k-1 +...+c1×s+c0=0; Among them, K p is the proportional gain, K d is the differential gain; T i is the integration time; T d is the differential time; s is a complex variable, b0, b1, ..., b m-1 , b m and a0, a1, ..., a n-1 , a n All are coefficients; c i K p , K d 、T i 、T d 、b y 、a x function; i = 0, 1, ..., k-1, k; y = 0, 1, ..., m-1, m; x=0, 1,..., n-1, n.

10. The PID parameter debugging device based on AI calculation according to claim 9, characterized in that: The AI ​​controller constructs a Routh table based on the closed-loop system characteristic equation, including the following steps: Step 1: Ensure c k >0; if c k <0, all coefficients of the closed-loop system characteristic equation are multiplied by -1; Step 2: Construct the Routh table. The first row of the Routh table is the highest power coefficient c. k Start taking the even-numbered coefficients, which are equal to c k , c k-2 , c k-4 , ...; The second row of the Routh table is the coefficient c from the next highest power k-1 Start taking odd coefficients, equal to c k-1 , c k-3 , c k-5 , ...; Starting from the third row of the Routh table, the element of each row = (the first item of the previous row × the item of the next column of the two previous rows - the item of the next column of the previous row × the first item of the two previous rows) / the first item of the previous row; If the first term in a row is 0 but the rest are non-zero, replace 0 with a very small positive number ∈ and continue the calculation; if a row is all 0, use the previous row to construct an auxiliary equation and take the derivative, and fill in the derivative coefficients in the new row.

Images